MicroRNA-Based Biomarkers in Pain

The journey of miRNAs to the diagnostic lab is still long and needs improved approaches at multiple levels (Angilini et al., 2015). The investigation of circulating miRNA as novel biomarkers for physiological and pathological conditions, owing to its non-invasive nature, can be both beneficial and cost effective. The selective sorting and excretion of miRNAs via lipoproteins or micro-vesicles, functioning as a mode of intercellular communication, is a fairly recent idea of crucial importance in relation to the potential of miRNA as biomarkers ( W. Stoorvogel, 2015 ). The advent of new and improved procedures for the purification and quantification of circulating miRNAs, especially quantitative RT-PCR, have paved way to enable the detection of low abundance miRNAs. Lack of reliable normalizers that can be used as universal standards is to be expected considering the dynamic nature of the composition of circulating RNAs. It is still necessary to standardize the profiling platforms, methodologies, including sample collection, storage, data normalization and statistical analysis, as well as the clinical parameters, before an miRNA can be developed into a reliable biomarker that can be used routinely. It has been suggested that miRNA clusters may render more specificity than an individual miRNAs (Angilini et al., 2015). The To capture biological variability, a central database cataloging the levels of miRNAs in both the healthy male and female populations, stratified by ethnicity, age, and geography, would serve as a useful reference and would greatly reduce the challenges associated with data reproducibility. Most of the studies published to date were performed on limited cohort size. The array of fluids and samples to profile just from blood include serum, plasma, PBMC, whole blood and exosomes. As mentioned earlier the results differ depending on the source (whole blood vs exosomes for CRPS ( Marguerite K. McDonald et al., 2014 ; Orlova et al., 2011 ), serum vs PBMC for fibromyalgia ( J. Bjersing et al., 2015 ; Cerdá-Olmedo et al., 2015 )). Data collection from selected patient populations along with controls can serve as a starting reference and can be tracked temporally to capture the effects of chronic pain, medical interventions and/or life style changes in these individuals. Identification of patients likely to experience pain relief without side effects, and predicting the risk of drug abuse prior to initiating therapy, though highly desired, is a huge unmet need. Combining cytokine and chemokine data with miRNAs will undoubtedly be beneficial in determining treatment options and for better understanding of the disease. Although pain is the common denominator in a variety of disorders, the underlying neurobiological mechanisms leading to pain can differ. Thus it is reasonable to expect an overlap of miRNA populations altered in different pathologies associated with pain, but every disorder, including different kinds of pain itself, may have its own unique miRNA signature. A number of studies investigating circulating miRNAs demonstrate their feasibility as potential biomarkers. Though the mechanisms that determine the loading of exosomes and the means by which they reach the recipient cells at distant locations though the circulation are not completely known, exosome biology is another aspect of circulating miRNAs that will undoubtedly open novel avenues in understanding normal physiology, disease mechanisms, and therapeutic intervention strategies. Accurate and precise diagnostic tests that are specific and sensitive enough to categorize patients with potential benefit from targeted therapies will be crucial for the success of personalized medicine. Clinical trials should be conducted on selected patient populations possessing molecular markers associated with specific drug targets. Ultimately, novel products and approaches successfully incorporated into clinical practice would evaluate our progress in making personalized medicine a reality ( Hamburg & Collins, 2010 ).

Chronic pain affects millions of people worldwide impacting the quality of life and posing enormous burden on healthcare systems (Gereau Iv et al.). Chronic pain includes neuropathic pain resulting from injury and malfunction of the nervous system, lower back pain, migraines, osteoarthritis etc. Traditionally, physiologic and behavioral responses are used to assess pain in animals and people ( Peterson & Servinsky, 2007 ). Pain scoring and assessment in humans are subjective and necessitates the development of objective measures to assess pain ( Chizh, Greenspan, Casey, Nemenov, & Treede, 2008 ). Development of new analgesics is stymied by high failure rates of drugs targeted to treat neuropathic pain. Thus, there is an urgent need to develop biomarkers that may potentially reduce failures in drug discovery ( Kola & Landis, 2004 ; Woodcock, Witter, & Dionne, 2007 ). Much of the focus in biomarker discovery in the pain field has been on secreted inflammatory mediators ( Marchi, Vellucci, Mameli, Rita Piredda, & Finco, 2009 ). Unlike cancer biology, where researchers have access to tissue samples, CNS disorders including pain are dependent on biological markers that can be identified from bodily fluids, predominantly blood. Small non-protein coding endogenous ~22 nucleotide RNA molecules called microRNAs (miRNAs) have attracted considerable attention in recent years both as potential biomarkers as well as for their utility in dissecting the molecular changes underlying various diseases. miRNAs play important roles in the regulation of gene expression and function by binding predominantly to the 3’ untranslated region (3’UTR) of target mRNAs that, in turn, causes cleavage or repression of translation of these mRNAs. Each miRNA species regulates multiple genes, and most mRNA targets contain multiple miRNA binding sites within their 3’UTR, rendering a complex regulatory network ( Bartel, 2009 ). Owing to their aberrant expression in a variety of human pathologies, miRNAs are often employed in the pursuit of biomarker identification and to discover and validate novel therapeutic targets ( Erson & Petty, 2008 ). The first reports of the presence of stable miRNAs in blood were published in 2008 ( X. Chen et al., 2008 ; Gilad et al., 2008 ; Mitchell et al., 2008 ). These reports were followed by a number of studies, predominantly in cancer biology, demonstrating the utility of miRNAs as biomarkers of the disease state and progression ( De Guire et al., 2013 ). It is now well established that miRNAs are present in the serum and plasma of humans and other mammals. This finding opens up the feasibility of using miRNAs as biomarkers of disease. Though the stability of miRNAs in serum was the initial concern, it has now been demonstrated that these circulating miRNAs are protected from plasma RNase activity and are, in fact quite stable ( X. Chen et al., 2008 ). The modulation of eukaryotic gene expression is regulated at various levels which include the basic genetic mechanisms, like initiation of transcription, transcriptional processing, post translational modification, and extends to epigenetic mechanisms. In addition to DNA methylation and histone modification mediated epigenetic modifications that influence access to DNA and thus protein expression ( Jirtle & Skinner, 2007 ; Vaissiere, Sawan, & Herceg, 2008 ), RNA-mediated gene silencing is another mechanism that is crucial in posttranscriptional gene silencing ( Bartel, 2009 ). miRNAs comprise a group of short non-coding single-stranded RNA molecules and potentiate epigenetic regulation of a significant proportion of the genome at the post transcriptional level. miRNA binds to the 8–10 nucleotide long, conserved ‘‘seed sequence’’ present within the UTRs of a target mRNA and thereupon exerts its effect. Numerous studies suggest the involvement of miRNAs in key biological processes such as development, immune system and cellular homeostasis, and their altered expression and regulation is observed in various pathological conditions such as autoimmune diseases, neurological and vascular diseases, infection and cancer ( Croce, 2009 ). In humans, genes coding for miRNA are known to be present on all chromosomes, with the exception of the Y chromosome and about 50% of them are located within the intergenic regions and possess their own regulatory elements ( Corcoran et al., 2009 ; Lagos-Quintana, Rauhut, Lendeckel, & Tuschl, 2001 ). The biogenesis of a miRNA is initiated in the nucleus with the transcription from its corresponding gene by either RNA polymerase II or RNA polymerase III, giving rise to a long (2–4 kb) primary transcript known as pri-miRNA, which is then poly-adenylated ( Cai, Hagedorn, & Cullen, 2004 ). The pri-miRNA is then processed into an imperfectly base paired stem loop structure of about 70–100 nucleotides known as precursor miRNA (pre-miRNA) by a microprocessor complex consisting of specific double stranded ribonucleases called Drosha, Pasha and accessory proteins ( Denli, Tops, Plasterk, Ketting, & Hannon, 2004 ; Lee et al., 2003 ). A RanGTPase and exportin 5 dependent mechanism aids the translocation of this pre-miRNA to the cytoplasm where it is cleaved into a short (about 22 bp long), imperfectly paired, double stranded mature miRNA-miRNA* duplex by another double stranded endonuclease, known as Dicer and accessory protein TRBP ( Chendrimada et al., 2005 ; Yi, Qin, Macara, & Cullen, 2003 ). Further processing by an argonaute protein, Ago2, allows one of the complementary strands of the duplex, known as the guide-strand to be incorporated into the miRNA-induced silencing complex (RISC), while the miRNA* strand, also known as passenger strand, is often degraded. This asymmetric incorporation into the RISC is supposedly determined by which miRNA strand 5′ end is less tightly paired ( Khvorova, Reynolds, & Jayasena, 2003 ). A Drosha-independent machinery identified in mammals, include the biogenesis of mirtons, spliced from the introns of mRNA coding genes and exported to the cytoplasm and processed by Dicer ( Okamura, Hagen, Duan, Tyler, & Lai, 2007 ; Ruby et al., 2007 ). The RISC bound miRNA possesses complementary sequence to the regions within the 3′UTR of the target mRNAs, and is the effector complex regulating their post transcriptional expression by various mechanisms such as mRNA cleavage, mRNA destabilization, and inhibition of translation ( Sara Elramah, Marc Landry, & Alexandre Favereaux, 2014 ). While the mechanism of mRNA cleavage requires perfect base pairing (observed in metazoans), the latter two mechanisms involve imperfect base pairing and are more common in mammals. The location of miRNA binding sites (seed region) on the target mRNA plays a key role in their regulation, with the most effective seed regions occurring at the 3’UTRs, and in some cases could also be potentiated by similar regions within the coding region. This could be attributed to higher affinity of the translation complex in comparison to RISC, within the 5′UTR and ORF of the mRNA, thus displacing the silencing complex ( Bartel, 2009 ). While the composition of the seed region may differ, it always involves a conserved Watson–Crick pairing to the nucleotides at 5′ region of the miRNA, resulting in diverse affinities and inhibition properties. These include canonical sites with a 6- to 8-nt match, non-canonical additional pairing at the 3′ end of the miRNA. Additional factors have been shown to affect miRNA seed efficacy such as AU-rich sites, the position of the seed within the 3′UTR of the target-mRNA ( Grimson et al., 2007 ), with sites at the extremities of the 3′UTR being more accessible than those in the middle because long UTRs may form occlusive interactions, thus greatly reducing miRNA site accessibility. miRNAs modulate paracrine signaling by their release into circulation in various forms and processes. Often transported either within extracellular vesicles or as RNase resistant complexes through their association with lipoproteins (high density lipoproteins) or RNA binding proteins (Ago2), these miRNA entities collectively, but not limited to, are termed circulating miRNAs ( Boon & Vickers, 2013 ; Grasedieck et al., 2013 ). Although it is still unclear whether all circulating miRNAs are Ago bound and/or associated in exosomes, miRNAs found in plasma were reportedly resistant to endogenous RNases ( Mitchell et al., 2008 ; Turchinovich, Weiz, Langheinz, & Burwinkel, 2011 ). The identification of several types of noncoding RNAs in the circulation within extracellular vesicles ( S, Mager, Breakefield, & Wood, 2013 ) has generated tremendous interest both from a biomarker perspective and for their therapeutic potential ( Alvarez-Erviti et al., 2011 ). Intercellular communication is predominantly thought to be mediated either through direct cell–cell contact or through the transfer of secreted molecules like peptides, hormones and cytokines. More recently, a third mode of communication involving extracellular vesicles has been recognized to play an important role in this intercellular information transfer ( Kowal, Tkach, & Thery, 2014 ; Raposo & Stoorvogel, 2013 ; S et al., 2013 ). This latter method of transport of biomolecular cargo has gained enormous attention due to the vesicular contents, including but not limited to several species of RNA, proteins, and lipids. It has been recently quite well substantiated that uptake of these vesicles can impart functional consequences in the recipient cells ( Raposo & Stoorvogel, 2013 ). Exosomes are extracellular vesicles 30–100 nm in size and are derived from the endocytic compartment of the cell. Exosomes are generated by the inward budding of an endosomal membrane that results in the formation of multivesicular bodies ( Raiborg, Rusten, & Stenmark, 2003 ; W. Stoorvogel, Kleijmeer, Geuze, & Raposo, 2002 ), which in a specific secretory pathway can fuse with the plasma membrane, releasing the exosomes either to the adjacent cell or into the circulation. Exosomes are released by various cell types, reticulocytes, and platelets and are present in most bodily fluids ( Kalra et al., 2012 ). The protein and lipid composition of exosomes differs from that of other types of extracellular vesicles including apoptotic bodies. In fact, exosomes are enriched in certain proteins and lipids and they are used as markers to differentiate exosomes from other types of extracellular vesicles, organelles, or cellular debris ( Gyorgy et al., 2011 ; Record, Subra, Silvente-Poirot, & Poirot, 2011 ). Exosomal composition is reflective of the cell secreting them. Changes due to infection, activation, inflammation, or transformation (in tumor cells) will influence and alter the composition of the exosome and thus are the basis for the biological consequences upon its uptake as well as for its biomarker utility. However, not everything that is present in the parent cell is incorporated into the exosomes, suggesting that this well-regulated process is dynamically altered by signalling cues ( Willem Stoorvogel, 2012 ; Villarroya-Beltri, Baixauli, Gutierrez-Vazquez, Sanchez-Madrid, & Mittelbrunn, 2014 ). An emerging area of research is to use exosomes purified from various sources and disease states to evaluate alterations in gene expression in recipient cells. Several methods are currently employed for exosome purification and a recent study compared downstream RNA profiling using exosomes obtained from different isolation methods. They showed that very pure exosomes is a prerequisite to obtain reliable omics data and identify exosome-specific functions and biomarkers ( Van Deun et al., 2014 ). In a recent editorial, the International Society for Extracellular Vesicles has put forth a minimal set of biochemical, biophysical, and functional standards that should be used to attribute any specific biological cargo or functions to EVs ( Lotvall et al., 2014 ). The precise molecular mechanisms and mediators responsible for determining the exosomal contents are still unknown. Manipulation of the parent cells can influence the packaging of biomolecules, and because exosomes can cross the blood–brain barrier, artificial manipulation of exosomal content, whether loading with nucleic acids or small molecules, is being explored for therapeutic intervention, including in CNS disorders ( S et al., 2013 ). Exosomes derived from antigen-presenting cells have been found to be protective in painful conditions. Dendritic cell-derived exosomes suppress the onset of collagen-induced arthritis in mice and reduce the severity of established arthritis ( Kim et al., 2005 ). Macrophage-derived exosomes also attenuated thermal hyperalgesia associated with complete Freund’s adjuvant (CFA) induced inflammatory pain ( M. K. McDonald et al., 2014 ). Thus exosomes derived from antigen presenting cells are capable of eliciting therapeutically relevant immune responses. Preclinical studies exploiting the immunogenicity of dendritic cell-derived exosomes for anticancer therapy are ongoing ( Viaud et al., 2010 ). Release and uptake of exosomes between neurons have been demonstrated in vitro ( Chivet, Hemming, Pernet-Gallay, Fraboulet, & Sadoul, 2012 ). Secretion of exosomes in an activity-dependent manner could thus mediate the interneuronal transfer of information, allowing signalling across synapses, necessary for plasticity. Exosomes can carry pathogenic proteins linked to neurodegenerative disorders such as Parkinson, prion, and Alzheimer diseases, suggesting that exosomes might contribute to the spreading of the pathology throughout interconnected cortical areas ( Chivet et al., 2012 ). Exosomes achieve epigenetic regulation through the transportation of a large number of transcribed mRNA. Analysis of exosomal transcriptome of the naïve or lipopolysaccharide (LPS) stimulated RAW 264.7 cells by next generation sequencing revealed protein coding potential of many transcripts; in addition to a high representation of histone H1 subunits within exosomes from unstimulated cells ( Marguerite K. McDonald et al., 2014 ). It has also been shown that extracellular histone H1 is neurotoxic and drives a pro-inflammatory response in microglia ( Gilthorpe et al., 2013 ). This could imply mature miRNAs to modulate inflammation by regulating proinflammatory mRNA or protein levels, while pre-miRNAs mediate a second regulatory wave much later in the recipient cells, thus potentiating a temporal mode of epigenetic regulation. Thus, exosomes provide the benefit of both delivering proteins and miRNA which facilitate direct and immediate action, and translatable transcripts which enable regulation of gene expression during inflammation ( Marguerite K. McDonald et al., 2014 ). The mechanisms involved in sorting of different exosomal cargo, exosome release, transport, and finally culminating in uptake by recipient cells is not completely understood ( Willem Stoorvogel, 2012 ; Villarroya-Beltri et al., 2014 ). However, there is a need to explore novel strategies involving pharmacological interventions capable of either selectively inducing the secretion of a protective exosome pool or inhibit the uptake of pathogenic ones. Exosome biology, in due course of time will open novel avenues in contributing to physiology, pathology and eventually therapy. From a biomarker perspective, characterization of exosomal content offers a glimpse to biological underpinnings within the body and can be beneficial in elucidating the mechanistic basis of the disease, severity and stages of disease progression ( Ajit, 2015 ). Several methods are available for miRNA detection. The expression profiles of several miRNAs in parallel can be determined by microarray analysis, deep sequencing or real-time RT-PCR. Northern blotting and in situ hybridization can be used to determine the level of individual miRNAs in tissues and cells ( van Rooij, 2011 ). The RT-PCR approach relies on reverse transcription of miRNA to cDNA, followed by qPCR with real-time monitoring of accumulation of reaction products. Two different formats of miRNA profiling, the OpenArray platform and Taqman Low Density Array (TLDA) are offered by Applied Biosystems, Foster City, CA. OpenArray platform is commonly used for large-scale miRNA profiling in a high-throughput format ( Hudson et al., 2013 ). These are 3072-well microfluidic OpenArray plates containing nano-liter-scale wells with pre-plated dried TaqMan primers and probes for up to 758 miRNAs and controls. For TLDA microfluidic cards, there is the option of screening 758 miRNAs in two 384 well cards format (Card A and B) or it can be custom designed to examine a small set of miRNAs (for example, miRNAs detected in plasma) in a medium-throughput format. Multiple additional kits and platforms such as those offered by Qiagen (Germantown, MD), Exiqon (Woburn, MA) and nCounter analysis system (Nanostring Technologies, Seattle, WA) can also be employed for qPCR. The advantages of TaqMan miRNA assays include high specificity and sensitivity. Though the platforms mentioned above provide comprehensive coverage of Sanger miRBase v14, the disadvantage of profiling for known miRNAs over sequencing is that we cannot identify novel miRNAs. The methodological challenges in utilizing miRNAs as circulating biomarkers and key procedural aspects of miRNA isolation contributing to variability in miRNA assessment from body fluids have been covered before ( Moldovan et al., 2014 ). Several methods are used in data normalization to account for variability. These include mean, quantile, endogenous and discovered miRNAs. For calculating the mean normalization, the average (CT0) of all miRNAs is subtracted from each CT value. Quantile normalization is commonly used in microarray expression analysis and uses the ranking of the raw CT expression values on each plate. Vendors usually recommend endogenous control miRNAs, in which case only these miRNAs are used for calculation of CT0. Alternatively, one can identify endogenous controls appropriate for the current study using various stability criteria. Standard deviation of the CT values across all samples as the stability measure for each miRNA can be determined and the most stable miRNAs can be selected as endogenous controls ( Orlova et al., 2011 ). The mean of these values can then be used as CT0 to calculate the delta CT (ΔCT) values of the target miRNAs on each plate (ΔCT = CTtarget − CTreference where CTreference is the average CT value of the endogenous controls)( Schmittgen & Livak, 2008 ). Statistical significance of differential expression can be determined for ΔCT values by performing a two-tailed independent samples t-test. Fold change can be determined according to the formula FC = 2^−ΔΔCT. The ΔΔCT values are calculated by subtracting the ΔCT values of the control individuals from the ΔCT values of the patient samples. Upon the completion of miRNA profiling, hierarchical clustering can be performed and the results can be combined with a heat map of the log-transformed fold changes for the significant miRNAs. The first study to demonstrate a role for miRNAs in mediating pain employed a conditional knockout of the miRNA biogenesis enzyme Dicer, within a subset of sensory neurons ( Zhao et al., 2010 ) in mice, and interestingly, these animals had reduced levels of several transcripts associated with nociception and exhibited attenuated inflammatory pain. Several studies have examined alterations in miRNA expression in tissue samples procured from rodent models of inflammatory and neuropathic pain models and have been reviewed ( K. K. Bali & Kuner, 2014 ; S. Elramah, M. Landry, & A. Favereaux, 2014 ; Kress et al., 2013 ; K. L. Kynast, Russe, Geisslinger, & Niederberger, 2013 ; McDonald & Ajit, 2015 ; A. Sakai & Suzuki, 2014 ; Tan, Pao, Cheng, Hung, & Liu, 2013 ). Several studies indicate roles of miRNAs including miR-103, miR-124a, miR-203 and miR-7a in mediating pain ( Favereaux et al., 2011 ; K. L. Kynast, Russe, Moser, Geisslinger, & Niederberger, 2013 ; A. Sakai et al., 2013 ; Sun et al., 2012 ). One of the first studies investigating circulating miRNAs in a chronic neuropathic pain condition was in whole blood from patients with complex regional pain syndrome (CRPS) ( Orlova et al., 2011 ). CRPS is a chronic neuropathic pain syndrome characterized by pain, inflammation, and aberrant sensory motor and trophic disturbances. The pathophysiology of CRPS is not completely understood, but its complex multifactor pathogenesis includes inflammatory, vascular, sympathetic nervous system, cortical, and spinal mechanisms. This study demonstrated the utility of circulating miRNAs in patient stratification. Of the three different groups that emerged from miRNA profiling, one group comprised 60% of CRPS patients; miRNA profiles from the remaining patients were interspersed among control samples from the other two groups. These findings suggest that clinically relevant patient stratification is possible on the basis of alterations in miRNA expression. Eighteen miRNAs were significantly different in CRPS patients. Analysis of inflammatory markers showed that vascular endothelial growth factor, interleukin-1 receptor antagonist, and monocyte chemotactic protein-1 were significantly elevated in CRPS patients. Several other markers exhibited trends that did not reach significance. Analysis of the patients who were clustered according to their miRNA profile revealed correlations that were not significant in the total patient population, again indicating that miRNA profiles can be useful in grouping patients and in identifying additional biomarkers, the significance of which may be diminished when the patient population is considered overall. This is particularly important in CRPS, a syndrome with differing symptoms ( Orlova et al., 2011 ). To determine whether exosomal miRNAs reflect the miRNA signature in whole blood of patients with CRPS ( Orlova et al., 2011 ), the same group of researchers analyzed miRNAs in exosomes purified from the serum of 6 patients with CRPS and 6 healthy controls ( Marguerite K. McDonald et al., 2014 ). A total of 503 miRNAs were detected in at least one human serum–derived exosomal sample and 127 miRNAs were significantly different between CRPS and control-derived exosomes. Sixteen of the 18 miRNAs dysregulated in patients with CRPS from their previous study were detected in human serum–derived exosomes, but only 5 of these (miR-25–3p, miR-320B, miR-939, miR-126–3p, and RNU48) were significantly altered. Additionally, the exosomal miRNA signature differed in the directionality of changes compared with that of the whole blood. In the exosomal fraction of patient blood, miR-320B, miR-939, miR-126–3p, and RNU48 were significantly upregulated, whereas they were downregulated in whole blood. Thus, hsa-miR-25–3p was the only miRNA exhibiting the same trend in whole blood and exosomes in patients with CRPS ( Marguerite K. McDonald et al., 2014 ). Authors suggest that analysis of exosomes from a larger sample of patients is required to determine whether exosomal miRNA could be a useful biomarker or can serve as a secondary strategy for patient stratification. The fact that exosomes can be purified from stored serum samples suggests the feasibility of retrospective analysis of banked samples. In a study investigating the relationship among pain, psychological characteristics, plasma cytokines, and miRNAs in whole blood from 22 healthy controls, 33 subjects with chronic pelvic pain (vestibulodynia, VBD) and 23 subjects with VBD and irritable bowel syndrome (IBS), the utility of miRNAs in patient stratification and for individual-based diagnosis and treatment was demonstrated ( Ciszek et al., 2015 ). They identified separate pathways for localized vs widespread pain predicted based on miRNA profiles. miRNA dysregulation in VBD is predicted to affect estrogen-relevant pathways and the authors suggest that this may be a contributing factor in localized pain to the pelvis. In patients with both VBD and IBS, miRNA alterations were linked to alterations in muscle, nerve, and glial cell function, thereby contributing to widespread pain. MicroRNA-132 modulates cholinergic signaling and inflammation in human inflammatory bowel disease. Higher levels of miR-132, with a correspondingly lower AChE (acetyl cholinesterase) activity were observed in IBD patients in comparison to healthy controls which support an inflammation-dependent homeostatic role in cholinergic signaling ( Maharshak et al., 2013 ). Aberrant miRNA expression has been identified in both bodily fluids and tissue biopsy samples from patients with other painful conditions. In patients with bladder pain syndrome, miR-199a-5p has been suggested to play a role in regulation of urothelial permeability,( Monastyrskaya et al., 2013 ) and expression of miR-449b and miR-500 in the bladder smooth muscle cells is increased ( Sanchez Freire et al., 2010 ). miR-146a was upregulated in different cell types and tissues in patients with rheumatoid arthritis, a systemic autoimmune disorder characterized by the inflammation of synovial tissue ( Ceribelli, Nahid, Satoh, & Chan, 2011 ), as well as in osteoarthritis ( Li et al., 2011 ). Altered miRNA expression profiles have been reported for immune disorders such as rheumatoid arthritis ( Ammari, Jorgensen, & Apparailly, 2013 ) and systemic lupus erythematosus ( Shen, Liang, Tang, de Vries, & Tak, 2012 ), painful conditions such as irritable bowel syndrome ( Fourie et al., 2014 ), chronic bladder syndrome ( Gheinani, Burkhard, & Monastyrskaya, 2013 ), visceral pain ( J. Zhang & Banerjee, 2015 ) migraine ( Andersen, Duroux, & Gazerani, 2015 ; Tafuri et al., 2015 ) and endometriosis ( Ohlsson Teague, Print, & Hull, 2010 ) when compared with healthy controls. Profiling of cerebrospinal fluid in ten women with fibromyalgia and eight healthy control subjects showed that the expression of nine miRNAs was significantly lower in patients when compared with healthy controls ( J. L. Bjersing, Lundborg, Bokarewa, & Mannerkorpi, 2013 ). Two recent studies identified differential expression of eight miRNAs in fibromyalgia patients compared to healthy women. One study was conducted in serum samples ( J. Bjersing, Bokarewa, & Mannerkorpi, 2015 ) and the second study used peripheral blood mononuclear cells (PBMCs) ( Cerdá-Olmedo, Mena-Durán, Monsalve, & Oltra, 2015 ). The eight miRNAs identified differed between the two reports indicating that the source of miRNAs being investigated is crucial and may result in the identification of different subsets of miRNAs. miRNAs altered in different painful disorders indicates that each disease has a specific miRNA “signature” rendering specificity for miRNA biomarkers. Pain is a very heterogeneous phenomenon and different etiologies and comorbidities likely contribute to the differences in miRNAs altered in these disorders. Currently, miRNA profiling in several neuropathic conditions reported do not provide a strong evidence of any particular miRNA to be exclusively involved in a molecular feedback mechanism to alleviate or aggravate pain due to a paucity in follow-up mechanistic studies. However, a few reports as stated above, show the regulation of pro and antiinflammatory miRNAs in the pain-inflammatory axis. We have investigated the role of hsa-miR-939, the miRNA that ranked first and showed a 4.3-fold downregulation (p = 6.0E-06) in CRPS patients ( Orlova et al., 2011 ). Hsa-miR-939 is predicted to target several proinflammatory genes, including interleukin-6 (IL-6), vascular endothelial growth factor A (VEGFA), and tumor necrosis factor-α (TNFα), TNFα-induced protein 1 (TNFAIP1); nuclear factor-kB2 (NFkB2); and nitric oxide synthase 2 (NOS2A). Our unpublished results suggest that hsa-miR-939 can regulate multiple proinflammatory genes and that downregulation of hsa-miR-939 in CRPS patients may increase expression of target genes, including those involved in amplification of an inflammatory pain signal transduction cascade. Studies focusing on individual miRNAs and their target genes, miRNA based patient stratification coupled to next generation sequencing analysis for potential miRNA target genes and pathway analysis are all strategies that could be beneficial in elucidating the functional significance of miRNA alterations in observed in various chronic pain conditions. Changes are needed on how new analgesics are discovered and developed because of the high rate of failure of drugs targeted to treat neuropathic pain ( Kola & Landis, 2004 ; Woodcock et al., 2007 ). Studies targeting well-defined patient populations for clinical trials and delineating placebo response will help in developing drugs efficacious in a subset of patients. Circulating miRNAs can potentially be used in patient stratification. This combined with the ability to predict treatment response would be immensely useful in the development of pain therapeutics. Development of experimental paradigms more predictive of trial outcomes ( Kola & Landis, 2004 ; Woodcock et al., 2007 ) are urgently needed. The ability to predict treatment outcome by a simple qPCR based blood test could revolutionize both treatment and clinical trials design. Efficacy trials could be conducted in mechanistically defined patient groups. Existing clinical trial samples can be used to conduct proof-of-concept studies to evaluate the feasibility and utility of assessing miRNA signatures. This will be beneficial in assessing the predictive validity of this approach. Assessing miRNA changes before and after treatment will provide insight into miRNA signatures and its alterations in good responders, non-treatment (placebo) responders and poor responders. Samples from both unsuccessful and successful clinical trials can provide meaningful insights and guidance for future or even ongoing trials. A patient’s miRNA signature may differ even prior to treatment initiation and changes in signature may correlate with treatment outcome. Identifying the potential target genes for individual miRNAs could provide insights into the molecular bases of different types of pain and help identify novel targets for pain therapy. If implemented successfully, these studies can potentially reduce future failures by targeting well defined patient population for clinical trials along with delineating placebo response. Ketamine, an N-methyl D-aspartate (NMDA) receptor antagonist a widely used anesthetic, is currently being studied in the treatment of depression ( aan het Rot, Zarate Jr, Charney, & Mathew, 2012 ) and pain, including CRPS. NMDA receptors are upregulated in CRPS causing the pain, allodynia and hyperalgesia that hallmark the disease ( Costigan, Scholz, & Woolf, 2009 ). Although controversial for its potential abuses and side effects (psychotomimetic effects), ketamine, when administered intravenously over several days with antipsychotics and antihallucinogenics, has been shown to provide relief to CRPS sufferers ( Borsook, 2009 ) that often lasts for weeks, or even months after treatment ( Dahanl et al., 2011 ; Schwartzman et al., 2009 ). Patients for whom all other treatments have failed are typically considered for ketamine treatment. Though ketamine can reduce pain, approximately 30% of the patients do not respond to ketamine therapy ( Dahan et al., 2011 ). There is no known clinical symptom or biomarker for predicting response to ketamine. In a recent proof-of principle study, intravenous ketamine treatment induced miRNA alterations in CRPS patients were investigated by profiling blood samples before and after treatment ( Douglas et al., 2015 ). The patients were grouped as good and poor responders based on whether they reported pain relief after ketamine therapy. Analysis of circulating miRNAs in CRPS patients undergoing ketamine therapy demonstrated differential expression of miRNAs in responders and poor responders before and after ketamine treatment. Differential expression of several miRNAs between good and poor responders before treatment indicates that miRNA profile can predict treatment response. The miRNA changes were not identical between responders and nonresponders, indicating that ketamine elicited response differed between the two groups ( Douglas et al., 2015 ). This study also investigated the mechanistic basis of miRNA alterations and provided molecular evidence suggesting that downregulation of hsa-miR-548d-5p in poor responders can influence the expression of its target mRNA UDP-glucuronosyltransferase UGT1A1. Authors propose that higher UDP-GT activity can lead to higher levels of inactive glucuronide conjugates and thereby can minimize the therapeutic efficacy of ketamine in poor responders ( Douglas et al., 2015 ). This is an example of how potential biomarker discovery can be extended to mechanistic studies that can generate insight on disease. Thus differences in miRNA signatures can provide molecular insights distinguishing responders from poor responders prior to treatment. Animal models have been crucial in understanding the mechanisms and maintenance of chronic pain and have an important role in facilitating translational research ( Gregory et al., 2013 ; Mogil, Davis, & Derbyshire, 2010 ). In conjunction with human studies, these models can be instrumental in elucidating the role of miRNAs in pain states. As mentioned above, recent reviews have cataloged the studies linking miRNAs to various pain conditions ( K. K. Bali & Kuner, 2014 ; S. Elramah et al., 2014 ; Kress et al., 2013 ; K. L. Kynast, Russe, Geisslinger, et al., 2013 ; Atsushi Sakai & Suzuki; Tan et al., 2013 ). Profiling studies have identified tissue-specific dysregulated expression of miRNAs in dorsal root ganglia (DRG) and spinal cord (SC) after peripheral inflammatory stimulus ( Bai, Ambalavanar, Wei, & Dessem, 2007 ; Kusuda et al., 2011 ), nerve injury ( Aldrich, Frakes, Kasuya, Hammond, & Kitamoto, 2009 ; K. Bali, Hackenberg, Lubin, Kuner, & Devor, 2014 ; Kusuda et al., 2011 ; A. Sakai et al., 2013 ; Strickland et al., 2011 ; von Schack et al., 2011 ; Yu et al., 2011 ) and cancer pain ( K. K. Bali et al., 2013 ). Additionally, differential regulation of individual miRNAs has been observed in specific regions of the brain in acute and chronic pain models ( Arai et al., 2013 ; Imai et al., 2011 ; Poh, Yeo, & Ong, 2011 ). The few studies that link specific miRNA changes to a pain condition suggest that temporal and spatial regulation of intracellular miRNA expression occurs at specific time points that correspond to early or late phases of pain conditions ( K. K. Bali & Kuner, 2014 ; K. L. Kynast, Russe, Geisslinger, et al., 2013 ; A. Sakai & Suzuki, 2014 ). miRNAs altered in animal models could be beneficial in elucidating the pathophysiology of pain; miRNA changes specific to certain models representing a pain condition could provide insights unique to the model being studied. While several recent studies have successfully established negative correlation between miRNAs and target mRNA levels ( H.-P. Chen et al., 2014 ; Im et al., 2012 ; Ni et al., 2013 ; Sengupta et al., 2013 ; Sun et al., 2012 ), a few have provided evidence for the role of specific miRNAs in impacting nociceptive behavior using rodent models ( Favereaux et al., 2011 ; K.L. Kynast, Russe, M√∂ser, Geisslinger, & Niederberger, 2012 ; A. Sakai et al., 2013 ; Willemen et al., 2012 ). miRNA induced attenuation or augmentation of pain phenotype has also been demonstrated in rodent models of inflammatory, neuropathic, and cancer pain ( K. K. Bali et al., 2013 ; Favereaux et al., 2011 ; K.L. Kynast et al., 2012 ; Pan et al., 2014 ; A. Sakai et al., 2013 ; Willemen et al., 2012 ). Circulating miRNAs can act extracellularly by directly interacting with cell-surface receptors. It was demonstrated that miR-21, miR-29a and let-7b can act as agonists of Toll-like receptors (TLR) 7 and 8 and thereby influence downstream signaling pathway in target cells ( Fabbri, Paone, Calore, Galli, & Croce, 2013 ). Secreted miRNAs represent a new class of pain mediators; extracellular let-7b is required and sufficient for inducing inflammatory pain via the activation of TLR7 and TRPA1. Let-7b is highly enriched in DRG and endogenous let-7b can be released from DRG neurons in an activity-dependent manner ( Park et al., 2014 ). This study suggests that miRNAs could potentially act in an autocrine fashion on nociceptors, or mediate novel signaling mechanisms. Profiling using blood obtained from multiple rodent models of pain, including spinal nerve ligation (SNL), spared nerve injury (SNI) surgical models of neuropathic pain and the administration of complete Freund’s Adjuvant (CFA) as a model of inflammatory pain showed differential expression of circulating miRNAs ( Qureshi et al., 2015 ). Circulating miRNAs differed between rats and mice indicating that there are species specific differences ( Qureshi et al., 2015 ). This study also investigated the temporal regulation of miRNA alterations in CFA model. Thus thirteen miRNAs were differentially expressed in the blood of mice three days after CFA-injection mice versus controls, while 31 miRNAs were significant in the blood ten days after CFA-injection. There were only three common differentially expressed miRNAs in the blood after three and ten days. The magnitude of downregulation increased for mir-384–5p, and mir-190b after ten days compared to changes observed after three days. An increase in number of miRNAs altered ten days after CFA-injection suggests that chronic inflammation can induce significant dysregulation of miRNAs. Thus different rodent models of pain have unique miRNA expression signatures in blood. Celecoxib is a selective COX2 inhibitor and extensively used to treat inflammatory pain including rheumatoid arthritis, osteoarthritis, headache, and musculoskeletal pain. Induction of COX-2 accounts largely for the high prostanoid production at the site of inflammation ( L. Chen, Yang, & Grosser, 2012 ). To study whether celecoxib administration can induce miRNA modulation, the drug was administered for 7 days in a CFA model of inflammatory pain. Comparison of four experimental conditions including saline, celecoxib, CFA, and CFA-Celecoxib co-administration showed that several miRNAs dysregulated by CFA were restored to baseline levels after the administration of celecoxib ( Qureshi et al., 2015 ). This study suggests that reversal of miRNA expression to physiological levels may be important for therapeutic effect of celecoxib. miRNA target prediction and pathway enrichment for rodent models of pain showed that Wnt signaling pathway was consistently enriched across all experimental conditions ( Qureshi et al., 2015 ). WNT family of secreted signaling molecules have an important role in both healthy and diseased states ( Rosso & Inestrosa, 2013 ). Binding of WNT ligands to Frizzled receptors can activate the canonical or β-catenin-dependent pathway, the noncanonical planar cell polarity pathway and the Wnt/Ca2+ pathway. Wnt signaling pathways are regulated by nociceptive input and their activation can mediate the development of acute and chronic pain ( Shi et al., 2012 ; Simonetti et al., 2014 ; Y. K. Zhang et al., 2013 ). Exosomes have a role in Wnt secretion and transport and their signaling activities have been functionally implicated in the development of Drosophila and in fibroblast-promoted cancer metastasis ( L. Zhang & Wrana, 2014 ).